1. Field
The present invention relates to communications in general and, in particular, to improving the transmission of information signals in a communications system.
2. Background
The quality of a communication link over a noisy channel depends on the energy to interference noise ratio Eb/No of the signal. To achieve a required bit error rate over the communication link, a particular Eb/No is required. The bit error rate is a function of several parameters including channel propagation characteristics. In order to reach the target Eb/No a transmitter must transmit a signal with sufficient power. In practice, communication systems of this type are power limited. In power limited systems the transmitter cannot necessarily transmit the amount of power required to maintain a desired bit error rate. In CDMA systems, the sum of the power required by each link in the system determines the overall capacity of the system. Thus, it is desirable for each communication link to require the lowest Eb/No possible.
In order to decrease the required Eb/No in CDMA systems, the data to be transmitted can be encoded. Many different encoders are known in the art. For example, conventional convolutional and turbo encoders are suitable for this purpose. All of the suitable encoders perform the same basic task of creating redundancy in the encoded information signal. In such encoding techniques, each encoded bit is a function of a plurality of input bits.
For example, the encoder system 1 of FIG. 1 can be used to provide a redundant encoded signal suitable for use in decreasing the required Eb/No in a CDMA communication system. The rate R encoder 4 of the encoder system 1 receives a stream of k information bits 2 and outputs a larger stream 6 of n coded bits wherein R is the code rate. The code rate R is the ratio of the number of information bits k per unit of time to the number of coded bits n per unit of time. Thus R=k/n, and n=k/R. The n bits of coded bit stream 6 at the output of the rate R encoder 4 can be transmitted over the transmission channel 8. A rate R decoder 12 performs a decoding operation that is the inverse of the operation performed by the rate R encoder 4. That is, the rate R decoder 12 converts the received n coded bits 10 into k information bits 14 that are substantially equivalent to the k information bits 2 that were input to the rate R encoder 4. In CDMA systems, typically the rate R=½ or R=⅓.
It is known that for similar encoding techniques a lower code rate R permits a lower Eb/No to obtain the same bit error rate (where it is understood that ⅓ is a “lower” rate than ½). However, this improvement in performance becomes negligible when the code rate R becomes too low. Typically little further improvement occurs below R=⅙. Furthermore, since the number of encoded bits increases as the code rate R gets smaller it is usually not desirable or even possible to transmit the large number of coded bits required for code rates lower than R=⅙. Typically, code rates of ½ and ⅓ are preferred.
Although the use of a lower code rate is desirable, because it would lower the required Eb/No in a CDMA communication system, it is deemed undesirable to use a lower code rate if doing so would have an overall adverse effect, such as lowering system capacity.
Lower code rates generate more bits for transmission than do higher code rates.
For example, if the code rate on a system were decreased from ½ to ¼, it would double the number of coded bits needed to be transmitted by the system. Thus, bandwidth between the remote station and the base station would need to be doubled in order to support such a decrease in code rates.
In a CDMA system, one could double the effective bandwidth on the forward link by halving the length of the Walsh codes used for orthogonally spreading the encoded bit stream. For example, by halving the length of the Walsh codes used in a CDMA system from 64 bits to 32 bits, a given data stream could be transmitted over the forward link in half the number of coded bits. Although decreasing the Walsh code length effectively increases the bandwidth between the remote station and the base station, it is undesirable to decrease the Walsh code length because doing so decreases the pool of Walsh codes. As is well known in the art, a decreased pool of Walsh codes decreases the number of users that the system can support. When a system has allocated all of its Walsh codes to users, no more users can be added to the system because the system is said to be “code limited”.
Since the number of spreading codes in a system is limited, the advantages of any gain achieved with a low code rate R can be offset by the disadvantage of the use of additional spreading codes. Thus, although decreasing the code rate R used by each user in a CDMA communication system improves the required Eb/No per user, it can also limit the number of users by creating a shortage of spreading codes. Although there exists ways of creating more spreading codes, such as by using quasi-orthogonal functions or by using multiple scrambling (PN) codes, these techniques are used as a last resort because they significantly increase the overall interference level in the system.
Besides being code limited, a system may be limited in the number of users it can support at a given time due to limits in the amount of power that the base station can transmit. Transmitting more power than is allowed will cause interference that cannot be tolerated by the adjacent cells. When a new user is added to the system, the amount of power that is transmitted by the base station will increase. Because there is a limit on the amount of power that the base station can transmit, the number of users may be limited by the total amount of power that can be transmitted. Therefore, even if there are additional spreading codes available, the number of users will be limited by the amount of power that can be transmitted by the base station. When a base station is limited in the number of users it can support at a given time due to power transmission limitations, the system is said to be “power limited.”
To improve the performance of a telecommunication system—performance which is usually measured in Erlangs, bits per seconds, or number of users—it is necessary to take into account both code limitations and power limitations. What is desired is a way to increase the system performance of a telecommunications system, often measured in the number of users that a telecommunications system can simultaneously support, by taking into account the fact that the system is both code limited and power limited.